Is there a stock bubble and will it burst soon?

DataSwarm avoiding the crash

NASDAQ index, from the dotcom crash in 2000 to September 2020 (Y Axis is log scale) – Graph courtesy Yahoo! Finance

We were asked by The Observer to comment on whether there is a bubble in the US stock market at the moment.  This post is a more detailed discussion. In essence, we believe there are 3 major considerations:

  • Traditional “end of bubble” signs are still in early days.
  • There are rational reasons for much of the rise in the market since the “Covid Crash” in February – March.
  • However, there are currently a large number of possible “bubble popping” events.

Taking these in turn:

Traditional End of Bubble signs

Looking at the chart above, there is clearly a bubble forming, but how close is it to bursting.

  • We cannot as yet see the typical “end of bubble” behaviours, i.e.: Large numbers of “ordinary” people being sucked into investing – though that is starting, with new financial trading apps offering free share dealing and owning fractions of shares in companies.
  • As yet the “Silicon Valley merely exists to channel money from Wall St to Madison Ave” effect has not yet appeared (i.e. the huge spends in PR and Marketing on popular media, where you can’t move but see discussions of stock trading).
  • Also, the recent 10% “market correction” has acted as a brake in the bubble’s development for a bit, the frothiest Tech stocks (even TSLA) have seen large corrections.

Some rational reasons for the rise

There are some rational reasons for explaining the rise so far of the leading stocks.

  • The rise has mainly been from companies that stand to profit hugely if there is a “phase shift” to a more digital, less physical world – Zoom, Microsoft, Netflix, Apple have all benefited hugely, as have the Covid drug/Healthcare companies whose shares have rocketed (Amazon gets a twofer deal – biggest global player in both home delivery from online ordering and B2B (B2C even) cloud platforms)
  • US Dollar has fallen c 10% from a pre-crash peak in February. This makes stock prices in US rise as many US Co’s have large revenues from overseas, so dollar values of that income (and profits) rise.
  • US Federal Reserve has poured money into US economy, which contributes to asset bubbles – and it’s not clear when this will end given the need to mitigate the economic effects of Covid.
  • Finally, most other investment options are worse for major investors trying to make money.

Beware of “bubble popping” events

The risk is more that some major event pushes the markets into a panic, much like the “Covid Crash” of Feb-March. And it’s hard to think of a time, post cold war, when there are so many very possible, very large shocks that could stop the developing bubble in its tracks and crash it, such as:

  • A very bad Covid Winter causes an even bigger Covid Crash.
  • Very bad US (and/or global) economic statistics raise the spectre of a major US/Global Depression. This would drive a major shift down in market sentiment.
  • There are a number of geopolitical powderkegs all over the world right now that could explode, and historically major plagues exacerbate any problems and drive major periods of instability in and of themselves.
  • Inconclusive US election + a very divided country could bring a period of extreme unrest, there are other predictive analysts predicting saying the US is in for extreme civil turbulence for a few years (and arguably it has started).

(In)conclusion – hedge your bets

Our experience of taking part in the IARPA 2nd  Geoforecasting Challenge last year (these are the “Superforecaster” challenges that Phillip Tetlock wrote about) was that it was useful to be specific about allocating “best efforts” odds to possible outcomes – and revising your view as circumstances change!

One of the hard parts of the IARPA challenges is that you have to predict the result on a particular day. In predicting the end of a bubble, timing is always the issue, even if you know it’s going to happen. Bubbles usually will carry on a lot longer then forecasters expect, and one can lose more money by exiting early than by seeing it to the fall. Thus, the way our predictive analytics system works is as much about risk analysis of the ease of getting out as an accurate prediction of when it will occur. How this works in practice is seen in the chart below.

DataSwarm avoiding the crash

Avoiding the crash – using DataSwarm predictive analytics in the stock market – DataSwarm vs S$P from Dec 2019

Not putting all eggs in a basket is one way of risk avoidance….but (as per the chart at the top of the page) if your systems are good enough, you can haul enough eggs out quickly enough – meaning you can leave the basket under the golden goose for longer.